INDEX

Explanations

repeated references to specific articles

oai_token-act-pair · gpt-4o-mini Triggered by @bot

but introducing contrast

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 redor

-0.40

供

-0.32

 planche

-0.31

lindungan

-0.30

 взаимодей

-0.30

]();

-0.30

 Vrij

-0.29

packageName

-0.29

equalsIgnoreCase

-0.28

 participe

-0.28

POSITIVE LOGITS

But

0.83

But

0.77

but

0.77

but

0.72

但她

0.72

SequentialGroup

0.69

但这

0.66

但我

0.65

但他

0.65

migrationBuilder

0.63

Activations Density 0.046%

repeated references to specific articles

but introducing contrast

No Comments

No Known Activations